November 1999 - Zope

Small improvement for Navigator
by Oleg Broytmann 15 Nov '99

15 Nov '99

Hi! Navigator.py, line 318: return renderer(self, PATH=self.PATH(), ^^^^^^ this allows passing of the entire namespace to renderer! Oleg. ---- Oleg Broytmann National Research Surgery Centre http://sun.med.ru/~phd/ Programmers don't die, they just GOSUB without RETURN.

1 0

RE: [Zope] ZSERVER, THREADS, PERFORMANCE & FUTURE PERFORMANCE
by Toby Dickenson 15 Nov '99

15 Nov '99

> -----Original Message----- > From: Jeff Rush [mailto:jrush@timecastle.net] > Subject: Re: [Zope] ZSERVER, THREADS, PERFORMANCE & FUTURE PERFORMANCE > > > On Thu, 11 Nov 1999 18:34:31 -0500, Michel Pelletier wrote: > > >However, the goal of threading Zope was NOT to increase requests > >per/sec, the goal was to allow simultaneous requests to > happen, in Zope > >1.x, requests were serialized and one requests wait for > (possibly long > >running) previous requests. Threading allows many requests to happen > >simultaneously, and means that individual threads can call 'blocking' > >methods without worrying about locking up the whole site. > > Related question -- if _adding_ threads doesn't gain much raw > performance, > then is it a sound workaround, on a site that (a) does a lot > of OODB writing > and (b) doesn't take a ton'o'hits, to turn threading off to > avoid the dreaded > ZODB conflicts. It seems this wouldn't lose much performance either. Today, the ODB maintains a separate pool of object-spaces (aka 'connections') for each version. The default size of the main pool is 7 object-spaces, and the default for version objects-spaces is 3. Assuming there are no versions in use, No more than 7 threads can be running at one time (even if you create more) since the excess will be blocked waiting for a free object-space. What if these write-intensive requests could specify that they should take place in a special pool, containing only one object-space..... The write-intensive requests would be serialized, but every other request would still be handled concurrently. ps. perhaps this should move to the zopedev list.

1 0

RE: [Zope] zope on NT problem
by Rob Page 15 Nov '99

15 Nov '99

> From time to time my zope server stop responding to http requests. > The python process is running, but any request to my site > doesnt produce a "200 ok" (or anything else) on the dos shell and the browser > request ends with a time out. Need a little more info: (a) What version of Zope and what version of NT (SP?) are you running? (b) Were you making requests from the machine on which the server was running? (c) What browser were you running? --Rob

1 0

Why can't I use propertyItems with propertysheets?
by Itamar Shtull-Trauring 15 Nov '99

15 Nov '99

I want to to call propertyItems and propertyValues in propertsheets, like <dtml-var "propertysheets.info.propertyItems">, but for some reason it won't work (2.1.0b1 - NotFound error). Is this a bug or a feature, and how can I solve it? The reason I ask is casue I'm trying to add a propertyDictinary method, which would be useful for doing things like: <dtml-call "manage_chnageProperties(anObject.propertysheets.info.propertyDictionary())"> i.e. copying all properties from one object to another easily. -- Itamar - itamars(a)ibm.net ---------------------------o----------------------------------------------o Perl/Gimp Greeting Cards | Trust? Ha! The US dollar is backed by ICBMs! | http://www.sealingwax.com | --Anonymous Coward, Slashdot |

1 0

Copy / Paste in DTML
by Itamar Shtull-Trauring 15 Nov '99

15 Nov '99

How can I copy and paste objects using DTML and/or PythonMethods? Passing the results of copy to the paste method (in python) doesn't seem to work, contrary to what the docs say. -- Itamar - itamars(a)ibm.net ---------------------------o----------------------------------------------o Perl/Gimp Greeting Cards | Trust? Ha! The US dollar is backed by ICBMs! | http://www.sealingwax.com | --Anonymous Coward, Slashdot |

1 0

propertyItems and in
by Oscar Picasso 15 Nov '99

15 Nov '99

In the ZQR I read: propertyItems Returns a list of (id, property) tuples. So in a DTML Document I tried: <dtml-in "propertyItems()[6:32]"> <dtml-var sequence-item> </dtml-in> It return the list of the actual properties. I thought that sequence-item would return somthing like: (id, value) so I could access the id with sequence-item[0] and the value with sequence-item[1]. Where am I wrong? I need to render both the property id and the property value inside the same <dml-in ...> iteration. Your comments will be greatly appreciated. Monday I have to show a demo of my little Zope application for a more important project. I'm a little nervous... ****************************************************** Oscar Picasso picasso(a)videotron.ca ******************************************************

3 2

How do I use "or" (|) with <dtml-if>
by Harry Henry Gebel 15 Nov '99

15 Nov '99

I have the following code for setting up colors: <dtml-if BGCOLOR> <STYLE type="text/css"> BODY { background: <dtml-var BGCOLOR> } </STYLE> </dtml-if> <dtml-if VLINK> <STYLE type="text/css"> A:visited { color=<dtml-var VLINK> } </STYLE> </dtml-if> I would like to have something like this: <dtml-if BGCOLOR | VLINK> <STYLE type="text/css"> <dtml-if BGCOLOR> BODY { background: <dtml-var BGCOLOR> } </dtml-if> <dtml-if VLINK> A:visited { color=<dtml-var VLINK> } </dtml-if> </STYLE> </dtml-if> Since this would eliminate having an extra STYLE tag. When expanded to include all the various optional tags that could be included this could save alot of space. However, I can not figure out how to use an "or" with dtml-if. I have not got that good of a handle on using python expressions with dtml, I assume this will be necessary. Of course, I could just take the <STYLE></STYLE> tags outside of any dtml-if, but this would leave an empty STYLE element in alot of pages which is aesthetically displeasing to me, although that is what I will do if I can't figure out a better way. Can anyone help me with this problem? ---------------------------------------------------------------------- Harry Henry Gebel ICQ# 43675297 West Dover Hundred, Delaware ----------------------------------------------------------------------

4 3

zope on NT problem
by julio dinis 15 Nov '99

15 Nov '99

Hi all, I cant understand this %!@#@ I start zope in a dos shell to be able to see the "200 ok" messages which means everything its okay :) >From time to time my zope server stop responding to http requests. The python process is running, but any request to my site doesnt produce a "200 ok" (or anything else) on the dos shell and the browser request ends with a time out. I just cant figure out in wich state is the server at this moments, cause the process is there. My server is not running as pcgi on iis, its a win32 zope normal install. I dont think its a memory problem cause when this crashes happen the python process isnt at the big values that sometimes he reach. Then if I do a stop and start, everything is okay :-0 Is there any specific debug I can do to figure out this situation? Thanks for your time, Julio ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

2 1

Acquisition isn't always bandwidth and browser-cache friendly, advice needed
by Farzad Farid 14 Nov '99

14 Nov '99

Hi, I have recently started working with Zope and I am in the process of building an intranet information site with it. The Acquisition paradigm is very powerful and interesting but I have a few remarks about its efficiency bandwidth-wise My site will mostly be composed of internal corporate documents, classified into folder, subfolders etc. I want a corporate logo to appear in top of each page. Therefore I have created a "logo" object in the top "images" folder, and each document uses the tag: <dtml-var "images.logo"> Now the big trouble : suppose I have lots of documents and a deep tree of folder, the HTML tags I get in each page are: <img src="MY_SITE/Chapter1/images/logo"> <img src="MY_SITE/Chapter1/Subchapter1/images/logo"> <img src="MY_SITE/Chapter1/Subchapter2/images/logo"> <img src="MY_SITE/Chapter1/Subchapter3/images/logo"> <img src="MY_SITE/Chapter1/Subchapter3/Subsubchapter3/imags/logo"> <img src="MY_SITE/Chapter2/images/logo"> <img src="MY_SITE/Chapter2/Subchapter1/images/logo"> etc. The same document appears on the site with dozens of different names, is transfered over the network dozens of times and encumbers the browser memory and disk cache dozens of times! How can this problem be solved elegantly without completely putting aside the acquisition process? How can I do it in this particular case with images? Thanks -- Farzad FARID <farzy(a)via.ecp.fr> Ingénieur Informatique Libre Alcôve - http://www.alcove.fr/

6 10

Handling bad URLs
by Clarence Gardner 14 Nov '99

14 Nov '99

(A.S. I actually kludged up a fix to my problem in the Zope code, but I think it's an interesting question anyway. ) Well, I did bad :( I sent out a mailing to a large membership list asking them to visit my Zope site. The URL was of the form /a/b?id=98723 where the id number is different for each person and opaquely identifies their record in the database. My problem is that I had a typo in the message, and the URL I actually sent was /a/b/id=98723 Now everyone who visits gets a 'Resource not found' error, unsurprisingly, and sadly, since the URLs are all unique, I can't put in a page telling them how to correct it. Is there any concept of an ErrorDocument or something that I could use to trap all these visits? Thanks again -=-=-=-=-=-=-=-= Clarence Gardner Software Engineer NetLojix Communications, Inc. NASDAQ:NETX clarence(a)netlojix.com

3 2